REVIEW Assessment of sexual function/dysfunction via patient reported outcomes

Transcription

1 (2008) 20, & 2008 Nature Publishing Group All rights reserved /08 $ REVIEW via patient reported outcomes Center for Sexual Medicine, Sheppard Pratt Health System and Department of Psychiatry, Johns Hopkins University School of Medicine, Baltimore, MD, USA The recent recognition of the high prevalence of sexual dysfunctions and disorders in our society, along with the substantial investment of the pharmaceutical industry in the field of sexual functioning, has resulted in a significant expansion in the development of valid and reliable measures of sexual function/dysfunction. The instruments tend to be brief self-report inventories, typically requiring min of patient time for completion. Most measures were initially developed as screening and outcomes measures for use in clinical drug trials of new treatments for sexual dysfunction, but are beginning to see more widespread use in the clinic. All these instruments must adhere to recently prescribed rigorous guidelines set forth by the Food and Drug Administration, and have been demonstrated to be valid and reliable indicators of the status and quality of sexual functioning in both men and women. The constructs that form the framework of our diagnostic system for sexual dysfunctions are not amenable to direct physical measurement, so that currently they must be assessed via these self-report scales. Although not as precise as physical measures, these inventories do an admirable job of quantifying and registering sexual functioning status in a concise and reliable manner, and have become indispensable tools in our clinical and research programs. (2008) 20, 35 44; doi: /sj.ijir ; published online 16 August 2007 Keywords: sexual measures; patient reported outcomes; sexual inventories Introduction Correspondence: Dr, Center for Sexual Medicine, Sheppard Pratt Health System and Department of Psychiatry, Johns Hopkins University School of Medicine, Baltimore, MD 21285, USA. LDerogatis@sheppardpratt.org Received 18 April 2007; revised 4 June 2007; accepted 25 June 2007; published online 16 August 2007 Although the systematic assessment of human behaviors and capacities via psychological measurement can be traced back to the Chinese over 1000 years BC, the beginning of contemporary psychological measurement is an innovation that dates to the turn of the twentieth century. 1 During this period, Galton developed the first psychological questionnaire, 2 Cattell invented the term mental test, 3 Adolph Meyer developed the first psychiatric rating scale at Johns Hopkins 4 and Woodworth devised the first self-report symptom inventory at Columbia. 5 Psychological assessment of human sexual behavior is absent from this register of measurement innovations, very much in keeping with the Victorian conventions of the time. At the turn of the twenty-first century, the situation has changed dramatically. The recognition of a high prevalence of sexual disorders in the community, 6 coupled with the rapid expansion of research on pharmacologic treatments for sexual dysfunctions, has resulted in the development of numerous new measures of sexual functioning. Much of the work on instrument development in this field has occurred during the past decade, paralleling the growth of research on pharmacologic treatments for sexual dysfunction, and reflecting associated demands made by the Food and Drug Administration (FDA). Many new pharmacological agents for the treatment of sexual conditions are currently in clinical trials, and in the case of men, have already been marketed. The research fundamental to these novel developments demands valid and reliable measures of treatment outcome that are brief, relevant and easy to use. Ideally, such measures will also have utility in the clinic, enabling clinicians to identify and diagnose sexual dysfunctions more easily, utilizing the same

2 36 measures by which potential treatments are evaluated in clinical research. For these reasons, it is important that clinicians as well as researchers in the field have accurate, up-todate information concerning the outcomes measures by which treatments are established as efficacious. The current review is designed to inform readers about the contemporary measurement options available for this purpose. In addition, it is important to understand the role that the FDA has had in establishing evaluative standards for these measures. In February 2006, the agency published a guidance titled, Patient-Reported Outcomes Measures: Use in Medical Product Development to Support Labeling Claims. 7 While these guidelines do not advance any new or innovative psychometric principles, they codify in detail the expectations that the agency has regarding what they consider to be validated outcomes measures. It follows therefore that the recommendations embodied in this guidance will have a strong influence on the nature and construction of instruments developed for this purpose. Since the length of this review must adhere to certain constraints, the PRO instruments discussed here are limited to those that have an established validation profile. One criterion that was used was that the instrument must have at least two published validation studies to be included. These measures assess sexual functionality, and address corollary questions concerning the specific nature of dysfunctional states. There are numerous additional questionnaires in various stages of development and validation, but which are not as far along in the validation process as those reviewed here. Space limitations prohibit incorporating these additional instruments in the current review. For a wider ranging review of sexually related measures in general, the reader is referred to a more comprehensive source. 8 The nature of psychological measurement In contrast to physical measurement (such as the measurement of weight or distance), the primary function of psychological measurement is typically to provide valid and reliable operational definitions for hypothetical constructs that lack tangible characteristics. Sexual functioning, hypoactive sexual desire disorder (HSDD) and sexual satisfaction are all hypothetical constructs that must be operationally defined before they can be measured and subjected to scientific manipulation. The methods and techniques of psychometrics provide the mechanism whereby they are defined, quantified and transformed into scientific variables capable of being evaluated through scientific data analysis. However, because the foci of measurement are constructs rather than tangibles, underlying measurement scales are not as sophisticated or powerful as those utilized in the physical sciences. This fact has often been misconstrued to suggest that psychological measures are soft or less scientific than measures of physical variables, a disapprobation that is unwarranted. What is true about psychological measurement in contrast to the measurement of physical variables, is that because accompanying measurement scales are not as sophisticated, resulting errors of measurement are larger. This fact does not make psychological measurement any less scientific, it simply means that it is less precise. Nosology of sexual dysfunctions A central issue in the measurement of quality of sexual functioning resides in the diagnostic system (nosology) we currently use to define sexual dysfunctions. The reason for this is that most of our measures of sexual functioning are designed around diagnostic models. There are two predominant nosologic systems currently in place for the diagnosis of sexual disorders: the World Health Organization s International Classification of Diseases-10 9 and the Diagnostic and Statistical Manual of Mental Disorders-IV-TR of the American Psychiatric Association. 10 There have been subsequent attempts to refine the nosologic definitions of the most prominent sexual dysfunctions. In 1992 the National Institutes of Health sponsored a Consensus Development Conference on Erectile Disorder. 11 The conference proved to be quite productive in that it resulted in a new definition for erectile disorder, and disseminated explicit guidelines for its diagnosis. A similar consensus conference was convened to refine diagnostic criteria for female sexual dysfunctions, under the auspices of the American Foundation for Urologic Disease. 12 At this conference, four broad categories of dysfunction were delineated: HSDD, sexual arousal disorder, orgasmic disorder and sexual pain disorder. Criteria for diagnostic assignments were clarified, the high prevalence of comorbidity among the female dysfunctions was highlighted, and a new subtaxonomic system for arousal disorders was introduced. In addition, manifest personal distress was reiterated as a necessary condition for establishing any diagnoses, and steps to achieve its quantification and measurement were strongly recommended. Current systems are very much a product of Masters and Johnson s 13,14 and later Kaplan s 15 thinking concerning the sexual response cycle, which strongly emphasized sequenced and coordinated phases of desire, arousal, orgasm and resolution. In this model, derangement or dysfunction can occur in any or multiple phases of the cycle, and the

3 diagnosis is usually made independent of a statement of etiology. The condition, so determined, is then assigned a label consistent with the dysfunctional phase (for example, HSDD and female arousal disorder). In keeping with this system, contemporary measuring instruments are designed to generate scores on domains comparable to the major phases of the sexual response cycle. They are intended to evaluate the respondent s perception of the quality of sexual activities during each phase, and the individual s satisfaction with these experiences. This mapping of outcomes measurement domains directly to diagnostic construct categories has been recently identified as a potential multiplier of measurement error in situations where the diagnostic system is imprecise; 16 however, for the time being this approach remains standard methodology. In addition to domain scores, many instruments also generate an aggregate total score, which is designed to reflect the respondent s status on an overarching construct such as overall quality of sexual functioning. Primary applications for measures of sexual function Basically, there are two primary areas of application for psychological instruments in the assessment of sexual function: screening and outcomes measurement. In the former application, a brief test or inventory may be used as a screen for presumptive evidence of sexual dysfunction and/or its level of severity. Such screening can be employed in primary care practices to identify patients with occult sexual dysfunction. 17 They are also employed in many clinical trials as part of formal inclusion/ exclusion criteria, (for example, the patient will be excluded if she does not score X15 on the Female Sexual Distress Scale (FSDS)). The screening functions of these instruments are many and varied, and may arise in both clinical and research applications. For an in-depth review of screening with psychological measures see DeRogatis and Culpepper. 18 In the case of clinical research, more often than not, PRO instruments measuring sexual function are used as primary and secondary outcomes measures. In this role, they are utilized to evaluate the efficacy of one or another form of treatment intervention, often a therapeutic drug. Since research with investigational drugs done in support of a New Drug Application (NDA) is closely regulated by the FDA, the standards put forward in the agency s recent guidance 7 strongly influence PRO design. To support a labeling claim tied to an effectiveness end point, the designated instrument must be demonstrated to be an instrument validated to establish such a benefit in the specific population under study. Essential characteristics of validated measures of sexual function As is true of scientific measurement in general, the two most fundamental characteristics of good psychological measurement are reliability and validity. The former refers to the consistency or replicability of measurement; it represents the ratio of true variation in the measurement to variation due to error. In contrast, validity addresses the essence of what is being measured; it reflects the degree to which an instrument measures what it purports to measure. Unlike reliability, which is established through a specifically prescribed series of studies, the validation of a measuring instrument is programmatic in nature. Validation is a continuing process that accumulates evidence from numerous exercises, studies and trials; it is enduring, at least theoretically, and validation operations serve to test continually and extend the generalizability of an instrument s validation envelope. Nunnally 19 has likened the validation process to an expanding network of circumstantial evidence supporting the validity of the test instrument. Table 1 lists the principal psychometric characteristics that are fundamental to validated PRO outcomes measures. Of these, discriminative validity is a relatively powerful form of validity, and two essential indicators of discriminative validity for measures of sexual function are sensitivity to functional versus dysfunctional status and sensitivity to therapeutically induced change. Both of these validity demonstrations are examples of criterionoriented validation, where test performance is validated against a logical external criterion (for example, presence of the condition or response to treatment). The former references an instrument s capacity to discriminate sexually dysfunctional individuals from persons free of any sexual disorders (its sensitivity and specificity in epidemiological terms), whereas the latter refers to an instrument s sensitivity to treatment-induced change. It is worth noting that a demonstration of one form of validity does not necessarily guarantee the other, since both capacities are not always present in the same instrument. Table 1 Principal features of validated instruments (PROs) to measure sexual functioning Reliability (internal consistency) Reliability (test retest) Validity (sensitivity to functional versus dysfunctional status) Validity (sensitivity to therapeutically induced change) Validity (convergent with other validated instrument(s)) Comprehensive content validation (readability, understanding and relevance) Brevity Unique dimension/domain variance Unique item variance (nonredundancy within domains) 37

4 38 Table 1 lists a series of additional features of validated instruments that measure sexual function. Convergent validation and brevity are also important features, but since these characteristics are relatively self-explanatory, they will not be discussed further. An additional feature that has been strongly emphasized in FDA guidance has to do with content validation. The agency has placed strong emphasis on this aspect of instrument development to reinforce their belief that patient reported outcomes must have patient input to be truly valid, in addition to that of experts. This means that item content should be developed with patient participation through focus groups and so on, and be evaluated by patients in terms of relevance to their condition and understandability. As to the remaining characteristics appearing in Table 1, unique dimensional variance, refers to the degree and pattern of correlations among the primary dimensions of the instrument. From an ideal perspective, the optimum test would comprise of dimensions or domains with low substantive overlap with each other, but at least moderate correlations with the aggregate (total) test score. Such a test design minimizes measurement redundancy, and facilitates a unique measurement contribution from each dimension comprising the instrument. Unique item variance represents an analogous concept concerning relationships between items within domains. Each test item should contribute uniquely to the measurement focus, and not simply represent another way of asking the same question. Specific measures of sexual function The discussion that follows, in combination with the detailed information contained in Table 2, characterizes 10 contemporary instruments designed to measure quality of sexual function, and additionally, the FSDS. The latter is included because it has become something of a standard in assessing sexually related personal distress among women, a demonstration that must be achieved before a DSM-IV diagnosis can be assigned. All these instruments have been developed during the recent past, with the majority having been validated during the past decade. These instruments vary primarily in terms of their levels of comprehensiveness; however, all have performed well against established psychometric criteria and have sound empirical evidence of reliability and validity. The Arizona Sexual Experience Scale (ASEX) is a five-item self-report inventory using a six-point Likert scale method. It was designed by McGahuey and his colleagues 20 to provide a pertinent, expedient and minimally intrusive method to evaluate sexual dysfunction and changes in sexual function in patients taking psychotropic drugs. The ASEX represents sexual function in men and women, regardless of sexual orientation or relationship with a partner. It measures quality of functioning in terms of five questions, each representing one domain: drive, arousal, penile erection/vaginal lubrication, ability to reach orgasm and satisfaction from orgasm. These were selected as the domains most commonly impaired by psychotropic drugs, and are consistent with domains of sexual function described in the DSM-IV, ICD-10, and currently used inventories of sexual function. The ASEX is interpreted based upon a total score and/or assessment of scores on individual items. It takes less than 15 min to administer and score. Reliability coefficients for internal consistency and test retest forms are excellent, and initial favorable test results of concurrent, convergent and discriminative validity have been reported. No norms have been published to date. It currently demonstrates many of the desirable characteristics of psychological outcomes measures, particularly brevity coupled with ease of administration and accurate quantification of core elements of sexual dysfunction. The Center for Marital and Sexual Health Sexual Functioning Questionnaire (CMASH-SFQ) 21,22 is a self-report inventory that consists of 21 specific items assessing sexual behavior in men. To date, apart from the inventory for partners, there is no adaptation of the CMASH-SFQ for women, and the inventory has not been tested in homosexual men. The authors describe the scale as ya conditionspecific quality-of-life instrument that assesses sexual functioning in men from both the patients and the partners perspectivesy. There are baseline and follow-up versions of the CMASH-SFQ, which share a core of 13 items. The inventory has scales for four domains related to sexual functioning: frequency of sexual activity, quality of erections, quality of orgasm and sexual satisfaction. It features concurrent administration to the partner as well as the subject. The CMASH-SFQ does not generate a total aggregate score. Standard reliability evaluations of the inventory have produced coefficients in the acceptable range, and a series of validation exercises have been reported. The inventory has demonstrated sensitivity to differences in functional versus dysfunctional samples, as well as sensitivity to therapeutically induced change. The Derogatis Interview for Sexual Functioning (DISF/DISF-SR) is a coordinated set of brief matched instruments designed to provide an estimate of the quality of an individual s current sexual functioning. 23 The DISF is semi-structured interview comprised of 25 items and reflects quality of sexual functioning in a multi-domain format. The DISF-SR is a matching self-report inventory designed to accomplish the same goal in a patient self-report mode. There are gender-specific male and female versions of both the DISF and the DISF-SR. All instruments in the DISF series are designed to be

5 Table 2 Descriptive and psychometric properties of 10 contemporary measures of quality of sexual function Inventory name Modality/ gender No. of items Administration time Domains Reliability Discriminative validity Sensitivity/ Specificity IC (a) TRT (r) IRR Functional/ Therapeutically dysfunctional induced change Published norms Arizona sexual experience scale (ASEX) SR/male and female 5 o10 min Drive, arousal, penile erection/vaginal lubrication, orgasm, satisfaction YES YES 0.82/0.90 NO Center for Marital and Sexual Health Sexual Functioning questionnaire (CMASH-SFQ) SR/male and partner 21 o15 min Sexual frequency, sexual satisfaction, orgasm, erectile function Y Y YES YES NO Derogatis interview for sexual functioning (DISF-SR) CI and SR/male and female 25 o15 min Cognition, arousal, behavior, orgasm, drive/relationship, overall total score YES YES 0.89/0.75 YES Female sexual function index (FSFI) Index of premature ejaculation International index of erectile function (IIEF) Profile of female sexual function (PFSF) SR/female only min Desire, arousal, lubrication, orgasm, satisfaction, pain SR/male only 10 o10 min Sexual satisfaction, control, distress YES YES YES YES YES NO SR/male only 15 o15 min Erectile function, orgasm, YES YES 0.97/0.88 YES desire, intercourse satisfaction, overall total score SR/female only 37 o20 min Desire, arousal, orgasm, pleasure, concerns, responsiveness, self-image YES YES 0.86/0.93 NO Sexual function questionnaire (SFQ) Sexual interest and desire inventory Short scale to measure female SR/female only 26 o15 min Desire, arousal sensation, arousal lubrication, enjoyment, orgasm, dyspareunia, partner relationship, overall total score YES YES YES CI 13 o15 min Overall total score 0.90 YES YES NO SR/female only 9 o5 min Feelings for partner, sexual responsivity, sexual YES YES 0.79/0.79 YES 39

6 40 Table 2 (Continued) Published norms Domains Reliability Discriminative validity Sensitivity/ Specificity Administration time No. of items Inventory name Modality/ gender Therapeutically induced change IC (a) TRT (r) IRR Functional/ dysfunctional frequency, libido, dyspareunia, partner problems sexual functioning (SPEQ) YES YES 0.93/0.93 YES, cutoff score (s) SR/female only 12/13 o5 min Unidimensional scale measuring sexually related personal distress. R version has an additional desire item Female sexual distress scale (FSDS/FSDS-R) Abbreviations: CI, clinical interview; IRR, inter-rater reliability; SR, self report; TRT, test retest reliability. interpreted at three distinct levels: discrete items, functional domains and aggregate summary (total) score. DISF items are arranged into five primary domains of sexual functioning: sexual cognition/ fantasy, sexual arousal, sexual behavior/experience, orgasm and sexual drive/relationship. In addition, an aggregate DISF total score is computed which summarizes quality of sexual functioning across the five primary DISF domains. Both the DISF and the DISF-SR take approximately min to administer. Internal consistency reliabilities for measures of the DISF-SR are well within acceptable ranges, as are test retest temporal stability coefficients. Inter-rater reliability estimates for the DISF interview were also very good. The DISF/DISF-SR has demonstrated good discriminative validity and sensitivity to treatment-induced changes in both male and female versions, and gender-keyed actuarial norms (in terms of area T-scores) are available for all versions of the test. The DISF/ DISF-SR is currently available in 12 foreign languages. It is worth mentioning that a revised version of the self-report instrument, the DISF-II, is currently being validated, and will be available in the near future. The Female Sexual Functioning Index (FSFI) 24 is a brief (19 item) self-report inventory designed to measure the quality of female sexual functioning. The FSFI represents sexual functioning on five primary dimensions of sexuality plus an aggregate total score. The FSFI was initially validated on a clinically diagnosed sample of women with female sexual arousal disorder (FSAD). Subsequently, the validation statement was extended to include women with a primary clinical diagnosis of Inhibited Female Orgasm Disorder or HSDD. 25 Internal consistency reliability coefficients were reported to be all in the 0.90s in the former study and in the acceptable range in the latter. The discriminant validity (that is, patients versus controls) of the FSFI was clearly demonstrated in both studies, as was confirmation of divergent validity. The Index of Premature Ejaculation (IPE) is a recently developed self-report inventory that focuses on the subjective aspects of the premature ejaculation experience, serving as a compliment to the time measures based on intravaginal ejaculatory latency time (IELT). The IPE was developed in a number of major stages involving developing the item pool for the prototype, content analysis through patient interviews, and sequential quantitative psychometric analyses. The process resulted in reducing the 17-item prototype to a final 10-item version. 26 As part of the psychometric analysis, the IPE was subjected to factor analysis based upon a relatively small sample of 85 men. Initially, four factors, accounting for approximately 59% of the variance were observed; however, after subsequent evaluation, only three were retained: sexual satisfaction, control and distress. A slightly modified

7 principal components analysis was conducted on the 10-item version, and a three-factor solution was observed. The IPE has been shown to have very good discriminant validity (known groups) and good convergent validity, correlating well with IELT. The IPE also showed high reliabilities, with internal consistency (a coefficients ) ranging from 0.70 to 0.86 for the three domain scores and test retest reliabilities varying between 0.72 and The International Index of Erectile Function (IIEF) is a 15-item self-report inventory developed by Rosen and his colleagues 27 to provide a brief measure of erectile function and capacity. It has been frequently recommended as a primary end point in clinical trials of erectile dysfunction, and has become a standard in that regard. The IIEF was developed in conjunction with the clinical trial program for sildenafil (viagra), and has since served as a major end point in more than 50 clinical trials. 28 It has now been linguistically validated in over 32 languages. The IIEF represents quality of male sexual function in terms of five domain scores: erectile function, orgasmic function, sexual desire, sexual satisfaction and overall satisfaction. The IIEF does not possess a total score. Both internal consistency and test retest reliabilities are superior for the scale, and there is factor analytic confirmation of the principal domains. Sensitivity and specificity are also very good, and concurrent validation against other comparable measures has been shown. Discriminative validity has been well established in comparisons of functional versus dysfunctional samples, and sensitivity to therapeutic change has been robustly demonstrated within the context of clinical trials of sildenafil and other treatments for ED. More recently, a five-item brief form of the IIEF termed the Sexual Health Inventory for Men (SHIM) has been developed and validated, along with an ED severity scale. 28 The Personal Experiences Questionnaire (PEQ) is a questionnaire developed by Dennerstein and her colleagues 29 with a specific focus on assessing the sexual functioning of middle-aged and older women. The original version of the PEQ was a 19-item inventory, which reflects six major dimensions identified through factor analysis: feelings for partner, sexual responsivity, sexual frequency, libido, vaginal distress/dyspareunia and partner problems. Although internal consistency and test retest reliability coefficients were within the acceptable range for the most part, several coefficients were rather low. In an attempt to maximize brevity, the authors subjected the full PEQ to mathematical optimization analysis with a focus on providing the best subset of items to optimize selected metric criteria. The optimization analysis reduced the number of items in the PEQ to eight. Subsequently, an additional item on orgasm was added, bringing the total number of PEQ items to nine. The resulting scale, which in later work has sometimes been referred to as the Short PEQ (SPEQ), continues to reflect the original six dimensions; however, four of the dimensions are represented by only a single item. In a subsequent validation study, 30 the nine-item SPEQ revealed internal consistency reliabilities in the range of , and test retest coefficients of Concurrent validity was assessed in comparisons against the DISF, where comparable domain scores correlated from 0.59 to In a small discriminative validity exercise, designed to differentiate family planning clients from a heterogeneous patient group, an optimized cut score of p7 demonstrated both sensitivity and specificity to be 79%. In a further very small sample of 13 women receiving sex therapy, only the libido dimension revealed sensitivity to treatment effects. The Profile of Female Sexual Functioning (PFSF) is discussed below. 31,32 In the development of the PFSF, preliminary items and domains of sexual function were identified through interviews with surgically and naturally menopausal women in the United States and Europe. A subset of items was selected for translation and further analysis. The resulting instrument is a self-report inventory developed by Proctor & Gamble Pharmaceuticals Inc. (Mason, OH, USA) to serve as a major outcomes measure in their trials of Intrinsa, a new transdermal testosterone treatment system for women suffering from low sexual desire. Qualitative linguistic validation was conducted in women with HSDD and non-hsdd women in eight countries to ensure items would have the same meaning across languages. The resulting instrument was evaluated in 332 oophorectomized women with HSDD and 258 age-matched non-oophorectomized controls. Psychometric item analyses resulted in 37 items organized into seven domains (sexual desire, arousal, orgasm, sexual pleasure, sexual concerns, sexual responsiveness and sexual self-image), characterizing female sexual function in menopausal women with HSDD. Excellent reliability and validity of the domains of the PFSF was observed in all geographies tested. Statistically significant differences between oophorectomized women with low libido and control women were found for all domains and all geographies, providing strong evidence of discriminative validity for the PFSF. In addition, measures of test retest and internal consistency reliability were well within accepted limits. The PFSF is an instrument specifically designed for measurement of sexual desire in women with low libido. Robust psychometric properties have been established in a large number of geographies and languages, making it useful for assessing therapeutic change in multinational clinical trials. A brief form of the PFSF has also recently been developed. 33 The PFSF has historically been treated as a proprietary instrument by Proctor & Gamble Pharmaceuticals; however, it 41

8 42 has very recently been made available to individuals in the health-care professions. The Sexual Function Questionnaire (SFQ) 34 is a self-report questionnaire designed as an outcomes measure of female sexual function. It is comprised of 26 items reflecting all aspects of the sexual response cycle desire, arousal and orgasm as well as dyspareunia. The item content was generated from the aggregated responses of 82 women to a semistructured interview. Factor analysis yielded seven domains of female sexual function: desire, physical arousal sensation, physical arousal lubrication, enjoyment, orgasm, dyspareunia and partner relationship. The item content of the SFQ was reviewed by an external panel of clinicians with expertise in psychology, physiology, gynecology, physical medicine and the treatment of FSD. Internal consistency of the domains ranged from 0.79 to 0.91 for all domains except partner relationship, which was 0.65, and test retest reliability is in the acceptable range. Validation studies to date include (1) two 12-week randomized, controlled clinical trials evaluating sildenafil citrate treatment versus placebo in preand post-menopausal women with a diagnosis of FSD that included FSAD (n ¼ 781); and (2) two normative 4-week trials in age-matched pre- and postmenopausal women without symptoms of FSD (n ¼ 201). Women in these studies were aged years. Discriminant validity was demonstrated with a significant difference between baseline mean SFQ domain scores of women with FSD compared to women without FSD (Po0.0001). Sensitivity to therapeutically induced change over time was established using a global efficacy question in both clinical studies. Substantial significant difference was evident between positive and negative responders to this question. The Sexual Interest and Desire Inventory (SIDI) is a brief, clinician-administered rating scale focused on measuring severity and change in response to treatment of HSDD in premenopausal women. The SIDI was developed in part through item response theory (IRT) analysis, with an initial prototype of 17-items, 35 which has since been paired down to a 13-item rating scale. 36 Reports have shown the SIDI to have high internal consistency reliability (a ¼ 0.90), and has demonstrated discriminant validity (known groups) relative to cases versus non-cases of HSDD, and has shown good convergent validity with domains from the FSFI in the same study sample. The SIDI has a somewhat unique measurement format in that items address both intensity and frequency of sexual events. The FSDS 37 was developed in an attempt to quantify and measure the construct of sexually related personal distress. The presence of manifest or measurable distress in presenting patients has been a required criterion for the diagnosis of FSD in both the DSM-IV and ICD-10 systems; however, no operational mechanism for measuring distress has ever been provided previously. Additionally, in the FDA s most recent guidance on clinical drug trials in FSD, 35 the agency has made presence of manifest personal distress a requirement for patient inclusion in clinical trials. The FSDS is a 12-item, unidimensional self-report inventory designed to measure the construct of sexually related personal distress. Repeated factor analyses of clinical trial samples have established the factorial invariance of the 12-item representation of the test, and the fact that it is unidimensional in nature. The large majority of the item variance in the measure is accounted for by a single common factor. Discriminant validity studies have been done with two samples of women with HSDD, one surgical, the other natural, versus women with no sexual complaint. In the natural menopause group, a cut score of X15 resulted in sensitivity ¼ 0.81, specificity ¼ 0.93 and PVP ¼ In the surgical menopausal group sensitivity ¼ 0.93, specificity ¼ 0.93 and PVP ¼ 0.89 using the same cut score. Comprehensive reliability analyses have been accomplished on the FSDS in several phase III study samples. In an early pilot study (n ¼ 78) with a 1-week time window, the test retest coefficient was In one clinical drug trial (n ¼ 190), the internal consistency coefficient (a) was 0.93, while the test retest correlation was 0.83 over a 4-week period. In a separate non-drug study (n ¼ 145) involving surgically and naturally menopausal women with FSD, internal consistency coefficients ranged from 0.93 to 0.94, while test retest (4 weeks) ranged from 0.82 to In a discriminant validity study attempting to identify accurately naturally menopausal women with FSD (n ¼ 188), ROC analysis was used to establish an optimal cutoff between patients and age-matched normal controls. A score of X15 had a positive predictive value of 0.91, with sensitivity ¼ 0.81 and specificity Results in a comparable sample of surgically menopausal patients (using X15) had a PPV of 0.89, with both sensitivity and specificity of Sensitivity to therapeutically induced change was also dramatically demonstrated in a separate trial of an agent to treat female sexual arousal disorder; comparisons of baseline scores on the FSDS compared to those at final visit revealed very dramatic reductions in personal distress across the study period (Po0.001). The FSDS is a very brief (approximately 5 min), psychometrically sound PRO instrument that is very easy to use. It has shown excellent ability to discriminate between FSD patients and normals, as well as having shown sensitivity to therapeutically induced change. It has been used in HSDD samples and women suffering from FSAD, with numerous different pharmacologic agents.

9 Conclusion: current status of the field In this review I have tried to provide an overview of the PRO instruments currently available to measure and quantify the status of an individual s sexual functioning. Psychometrics has progressed dramatically in sexual medicine during the past decade to a field of measurement in which a great deal of new development is taking place. Much of the impetus for the growth of measurement in this field has come directly from a dramatic increase in the pharmaceutical industry s investment in the area of sexual medicine. Industry s interest is tied to the recent development of a variety of new compounds to treat sexual dysfunctions, and a corresponding FDA directive concerning the requirements for validated instruments to serve as outcomes measures in related clinical trials. As alluded to earlier, the PRO instruments covered in this review are primarily contemporary measures that have been developed during the past decade. The design of these measures tends to feature brevity and the self-report modality, which articulate well with the principal formats for clinical trials. The measurement constructs these inventories and clinical interviews are designed to operationalize tend to be derived from the constituent elements of the sexual response cycle and their derangement, and are referenced under the general headings of desire, arousal and orgasm. Some of the PRO instruments reviewed here are designed for only one gender, while other measures feature both male and female versions of the scale. Most of the instruments used in sexual medicine have been designed as self-report inventories; however, several scales have also been developed as brief interviews. The latter tend to be quite similar to their self-report counterparts in content (basically, identical item sets); however, the clinical interview format introduces professional clinical judgment and greater flexibility into the evaluation process. In the context of clinical trials, these instruments provide an important quantitative adjunct to the very basic event diaries that have historically been utilized as primary end points in treatment trials in this area. Because the diaries represent a very primitive measurement form (that is, counting or enumeration ), they are not highly sensitive to treatment effects unless the effect size under study is quite large. Psychometrically derived instruments provide a much more flexible and sophisticated evaluative mechanism to assess outcomes efficacy, and they do so in an efficient, straightforward and inexpensive manner. 38 Although the majority of the instruments discussed here have been developed within the context of clinical research, nothing precludes the interested clinician from utilizing them in clinical practice. By adopting these measures for clinical applications, clinicians could make significant contributions to the instruments validation processes, and by utilizing evaluative techniques in the clinic identical to those utilized to establish treatment efficacy in clinical research, sexual medicine s translational efforts would be substantially advanced. References 1 DuBois PH. A History of Psychological Testing. Allyn & Bacon: Boston, Galton F. Inquiries into Human Faculty and its Development. MacMillan: NY, Cattell JM. Mental tests and measurements. Mind 1890; 15: Kempf EJ. The behavior chart in mental diseases. Am J Insanity ; 7: Woodworth RS. Personal Data Sheet. Stoelting: Chicago, Laumann EO, Paik A, Rosen RC. Sexual dysfunction in the United States: prevalence and predictors. J Am Med Assoc 1999; 281: US Department of Health and Human Services, Food and Drug Administration, Center for Drug Evaluation and Research (CDER). February Guidance for industry patient reported outcomes measures: use in medical product development to support labeling claims. US Department of Health and Human Services, Food and Drug Administration, Center for Drug Evaluation and Research (CDER): Rockville, MD, Davis CM, Yarber WL, Bauserman R, Schreer G, Davis S (eds). Handbook of Sexuality-Related Measures. Sage Publications Inc.: Thousand Oaks, CA, World Health Organization. ICD-10: International Statistical Classification of Diseases and Related Health Problems. World Health Organization: Geneva, American Psychiatric Association. Diagnostic and Statistical Manual of Mental Disorders, 4th edn. American Psychiatric Association: WA. 11 NIH Consensus Development Panel on Impotence. Impotence. J Am Med Assoc 1993; 270: Basson R, Berman J, Burnett A, Derogatis L, Ferguson D, Fourcroy J et al. Report of the international consensus development conference on female sexual dysfunction: definitions and classifications. J Urol 2000; 163: Masters WH, Johnson VE. Human Sexual Response. Little Brown: Boston, Masters WH, Johnson VE. Human Sexual Inadequacy. Little Brown: Boston, Kaplan HS. Disorders of Sexual Desire. Brunner/Mazel: New York, Derogatis LR, Burnett AL. Key methodologic issues in sexual medicine research. J Sex Med 2007; 4: Utian WH, MacLean DB, Symonds T, Symons J, Somayji V, Sission M. A methodology study to validate a structured diagnostic method used to diagnose female sexual dysfunction and its subtypes in post-menopausal women. J Sex Marital Ther 2005; 31: Derogatis LR, Culpepper WJ. Psychological tests in screening for psychiatric disorders. In: Maruish M (ed). The Use of Psychological Testing for Outcomes Assessment and Treatment Planning. Lawrence Erlbaum: Mahwah, NJ, Nunnally J. Psychometric Theory. McGraw-Hill: NY, McGahuey CA, Gelenberg AJ, Laukes CA, Moreno FA, Delgado PL. The Arizona sexual experience scale (ASEX): reliability and validity. J Sex Marital Ther 2000; 26: Corty EW, Althof SE, Kurit DM. The reliability and validity of a sexual functioning questionnaire. J Sex Marital Ther 1996; 22: Glick HA, McCarron TJ, Althof SE, Corty EW, Willke RJ. Construction of scales for the Center for Marital and Sexual Health (CMASH) Sexual Functioning Questionnaire. J Sex Marital Ther 1997; 23:

10 44 23 Derogatis LR. The Derogatis interview for sexual functioning (DISF, DISF-SR): an introductory report. J Sex Marital Ther 1997; 23: Rosen R, Brown C, Heiman J, Leiblum S, Meston C, Shabsigh R et al. The female sexual function index (FSFI): a multidimensional self-report instrument for the assessment of female sexual function. J Sex Marital Ther 2000; 26: Meston CM. Validation of the female sexual function index (FSFI) in women with female orgasmic disorder and women with hypoactive sexual desire disorder. J Sex Marital Ther 2003; 29: Althof S, Rosen R, Symonds T, Mundayat R, May K, Abraham L. Development and validation of a new questionnaire to assess sexual satisfaction, control, and distress associated with premature ejaculation. J Sex Med 2006; 3: Rosen RC, Riley A, Wagner G, Osterloh IH, Kirkpatrick J, Mishra A. The international index of erectile function (IIEF): a multidimensional scale for assessment of erectile dysfunction. Urology 1997; 49: Rosen RC, Cappelleri JC, Gendrano III N. The international index of erectile function (IIEF): a state-of-the-science review. Int J Impot Res 2002; 14: Dennerstein L, Lehert P, Dudley E. Short scale to measure female sexuality: adapted from McCoy female sexuality questionnaire. J Sex Marital Ther 2001; 27: Dennerstein L, Anderson-Hunt M, Dudley E. Evaluation of a short scale to assess female sexual functioning. J Sex Marital Ther 2002; 28: Derogatis LR, Rust J, Golombok S, Davis S, Bouchard N, Rodenberg C et al. Validation of the profile of female sexual function (PFSF) in surgically and naturally menopausal women. J Sex Marital Ther 2004; 30: McHorney CA, Rust J, Golombok S, Davis S, Bouchard C, Brown C et al. Profiles of female sexual function: a patientbased, international, psychometric instrument for the assessment of hypoactive sexual desire in oophorectomized women. Menopause 2004; 11: Golombok S, Derogatis L, Rodenberg C, Koochaki P, Schmitt S, Rust J. Development and validation of a new screening tool for hypoactive sexual desire disorder: the brief profile of female sexual function r (B-PFSF r ). Gynecol Endocrinol 2007 (in press). 34 Quirk FH, Heiman JR, Rosen RC, Laan E, Smith MD, Boolell M. Development of a sexual function questionnaire for clinical trials of female sexual dysfunction. J Womens Health Gend Based Med 2002; 11: Sills T, Wunderlich G, Pyke R, Segraves RT, Leiblum S, Clayton A et al. The sexual interest and desire inventory female (SIDI-F): item response analysis of data from women diagnosed with hypoactive sexual desire disorder. J Sex Med 2005; 2: Clayton AH, Segraves RT, Leiblum S, Basson R, Pyke R, Cotton D et al. Relaibility and validity of the sexual interest and desire inventory-female (SIDI-F): a scale designed to measure severity of female hypoactive sexual desire disorder. J Sex Marital Ther 2006; 32: Derogatis LR, Rosen R, Leiblum S, Burnett A, Heiman J. The female sexual distress scale (FSDS): initial validation of a standardized scale for assessment of sexually related personal distress in women. J Sex Marital Ther 2002; 28: Althof S, Rosen R, Derogatis LR, Corrty E, Quirk F, Symonds T. Outcomes measurement in female sexual dysfunction clinical trials: review and recommendations. J Sex Marital Ther 2005; 31: